Inmunoprecipitaciones con sueros de pacientes con NMT

RF ray tracing simulations were performed by the authors to check the validity of the claim that the total channel delay spread may decrease as the distance from the BS increases. Fig. 4.4 shows a three dimensional view of the simulated hilly environment, which was chosen to be in Konya, Turkey. On the other hand, Fig. 4.5 depicts the project elevation view with the place of the transmitter and the sets of receiver locations. An isotropic transmitter with an input power of 45 dBm was placed above all the receivers with a center frequency of 100 MHz and a bandwidth of 5 MHz. The receivers sensitivity were set to −120 dBm. The receivers were divided into different sets, where each set is characterized by a constant planar distance from the transmitter. It was desired to deal with the worst case (largest delay spread), and therefore each set was represented by the receiver having the maximum delay spread within this set. The sets were then arranged according to their maximum delay spread and via simulations it was found that at distances above 6.5 km, the maximum delay spread of each set decreases as the distance of the set from the transmitter increases.

Chapter 4. Software-Based FM Wireless Channel Model 38

Figure 4.4: Three dimensional view of the simulated environment in Konya, Turkey.

Figure 4.5: Project view of performed Wireless Insite simulation in Konya, Turkey. Isotropic transmitter is placed in the center and receivers were placed as cylinders

around the transmitter.

For distances below 6.5 km, the simulation results were opposite of the earlier case. The physical interpretation for this is that at small distances there is a high chance of receiving LOS from the transmitter, which has very large power compared to the power of the rest of the clusters. This reduces the overall delay spread. This still does not mean that the receiver did not receive other clusters. However this effect will be more apparent in the maximum excess delay rather than the root-mean-square (rms) delay spread. This is because the maximum excess delay does not deal with the relative powers of different MPCs as long as the power of each of them is larger than the receiver sensitivity value. Moreover, that decrease in the total delay spread as the distance from the transmitter increases supports our proposal concerning the decrease in the number of clusters.

Chapter 4. Software-Based FM Wireless Channel Model 39

4.5

Chapter Conclusion

With the availability of more extended measurements, this model can be taken as a basis for new sub-models in order to better fit the FM Band. It is observed that one of the characteristics of the FM Band channels is the increase of the angular spread due to the increasing number of clusters. Hence, systems based on MIMO antenna technique can be deployed in the FM Band with high performance, resulting in a higher bits per hertz systems. Another key observation is the presence of very large delay spreads in the FM Band. For the deployment of systems based on orthogonal frequency division multiplexing (OFDM), this implies a very large Cyclic Prefix (CP). This leads to a very long symbol duration, which in turn means low data rate. Therefore, the utilization of generalized frequency division multiplexing (GFDM) considered for 5G systems might outperform OFDM systems since it can overcome the problem of the channel coherence bandwidth being smaller than the bandwidth of each sub-carrier [40].

Next chapter includes the implementation of a hardware channel emulation technique. The channel effects are introduced to the signal in the RF domain without the need to process the signal in the baseband after its downconversion.

Chapter 5

Hardware implementation

Typically, the received signal is the processed version of the transmitted signal. This processing is done by the channel where time and frequency dispersion are being intro- duced to the signal. Noise is then added to the smeared version of the signal. This noise is widely accepted to be modelled as Additive White Gaussian (AWG) that has a flat power spectrum over all frequency contents and whose distribution is Gaussian with zero mean. The addition of the noise to the smeared signal is a simple task where there is plenty of AWG noise (AWGN) generators that can be utilized. This chapter will show the proposed method to introduce the channel effects -namely time distortion and Doppler effect- onto the signal in the bandpass.

5.1

Channel delay taps

It was mentioned earlier that the approach of introducing delay to the channel taps by means of cables extension turned out to be impractical especially when a large delay is required. According to Table 3.1, it is clear that low frequencies of operation in the outdoor environments are accompanied with large delay spread values. Therefore, to emulate a wireless channel operating in a low frequency, the cables approach becomes non-feasible due to the bulkiness of the hardware emulator. The choice was to emulate a predefined outdoor channel operating at 140 MHz that is typically characterized by large delay spread values.

Chapter 5. Hardware implementation 41

A surface acoustic wave (SAW) is a type of mechanical wave motion traveling along the surface of a solid material. The wave was discovered in 1885 by Lord Rayleigh. A basic SAW device consists of two interdigital transducers (IDTs) on a piezoelectric substrate such as quartz as shown in Fig. 5.1 [7]. In a SAW filter, the electrical signal is converted to an acoustic wave and then back to an electrical signal. Large delays are easily acquired making use of the advantage that acoustic waves travel very slowly (typically 3000 m/s). The IDT geometry is capable of almost endless variation, leading to a wide variety of devices. Starting around 1970, SAW devices were developed for oscillators, and bandpass filters for domestic TV and professional radio. In the 1980s the rise of mobile radio, particularly for cellular telephones, caused a dramatic increase in demand for filters. New high-performance SAW filters are produced in large numbers.

Figure 5.1: Basic SAW Device [7].

The idea of utilizing SAW filter was adopted in [34]. According to the previous explana- tion, SAW devices can be designed to introduce large group delays that can be exploited to generate large multipath delays in laboratory environments. Note that most of the commercially available SAW filters are bandpass filters with different center frequencies. Therefore, for the case of large multipath channel delay taps being required, the use of SAW filters is preferred to the use of cables because they would be much smaller in size and cheaper than the bulky elongated cables. Moreover, many of the SAW filters manufacturers offer the luxury of ordering a customized SAW filter that gives the desired group delay values. This is in addition to the availability of a large variety of SAW filters offering different group delay values at the desired bandpass. Note that their phase response is almost linear. This results into very small tolerance in the specified group delay along the whole passband.

Chapter 5. Hardware implementation 42

In [34], two emulation techniques were described: The Cascade Connection Technique and the Feedback Technique. The cascade technique is simple where SAW filters op- erating at the same center frequency are being cascaded to offer a larger delay for the tap if required. The amplitude and the phase of the tap can be controlled via attenu- ator/amplifier and phase shifter respectively. Another choice is to replace the cascaded small-delay SAW filters with another one that operates at the same center frequency but offers a higher group delay. This solution is feasible due to the availability of a wide range of SAW filters at a certain frequency as mentioned before. The cascading technique is shown in Fig. 5.2.

Figure 5.2: Cascading technique of SAW filters

The feedback technique basically depends on entering the output of the first tap to the SAW filter once again in order to result into other taps that are equally spaced in time domain with the same group delay amount the SAW filter offers. The amplitude and the phase of the first tap is fully controllable through the attenuator/amplifier and the phase

Chapter 5. Hardware implementation 43

shifter respectively. As for the rest of the taps resulting from the loop, there is no control over their amplitudes and phases since they depend basically on the first entering data sample. However, the generic shape of the resulting taps amplitudes can be controlled via gain blocks (attenuators/amplifiers) where an exponentially decaying power response might result. Note that the SAW filter itself also has its own insertion loss, which is also in scale of dB. Thus, the resulting PDP is typically an exponential decay even when no attenuators/amplifiers are used. The number of the resulting taps from the loop can also be controlled through a variable attenuator that can play a role in determining the maximum excess delay. The feedback technique is shown in Fig. 5.3.

Figure 5.3: Feedback technique

The cascaded technique is helpful when the predefined channel has relatively low number of channel taps. In such a case the cascaded technique would be more efficient than the feedback technique due to the presence of full control over the amplitudes and phases of the taps, in addition to the possibility of getting unequally spaced taps in time if it is desired. However, in many cases especially when the total channel delay spread is large, the number of desired channel taps to be emulated is large and hence, it might be impractical and very expensive to assign each of those taps a dedicated branch as shown in Fig. 5.2. In that case, it may be a possible solution to utilize the feedback technique in emulating clusters. As mentioned before in Chapter 2, the PDP within each cluster generally decays exponentially with time. Therefore, the feedback technique can be used to emulate a single cluster where the first tap of the cluster is fully controlled while the others are not. Thus, the proposed emulation method when there is more than one cluster with large number of taps required per cluster is the combination of both of the techniques with each branch resulting into one cluster whose PDP decays exponentially with time. Each cluster is assigned to a branch where the maximum excess delay of the cluster can be controlled via the variable attenuator of the corresponding branch.

Chapter 5. Hardware implementation 44

Note that at low frequencies, there is typically more than one cluster in the outdoor environment. Fig. 5.4 shows the proposed emulation hybrid technique. There, the SAW filters on the left side of the figure are for determining the cluster’s first tap delay while the SAW filters in the loops are for determining the time delay between each two successive taps within the same cluster.

Figure 5.4: Hybrid technique with each branch corresponding to a cluster whose maximum excess delay can be controlled via the variable attenuator of that branch

Though its inability in controlling all the clusters taps, this method is still considered to be cost efficient when outdoor low frequency channels are desired to be emulated with a very large number of taps per cluster required. It is worth to remind that the number of required taps in the channel delay spread depends on the symbol duration of the utilized communication systems. For instance, the transmitted symbol duration in the modern OFDM systems is typically large, and thus large separation between the channel taps can be tolerated that will decrease the overall number of required taps. Note that the hardware can be reduced even more by uniting the feedback block of the emulator for all of the clusters as shown in Fig. 5.5, and thus getting rid of the feedback block accompanied with each cluster. This comes with the price of forcing all the clusters to have the same Variable Attenuation value. This does not offer high flexibility in

Chapter 5. Hardware implementation 45

controlling each cluster separately. This method isn’t recommended since the number of emulated clusters is typically not that large even for outdoor low frequency channels according to the reviewed literature. Thus, the feedback loops that are to be saved are not that effective as compared to the offered flexibility when each cluster has its own feedback loop where there would typically be a few increment in the hardware.

Figure 5.5: Illustrative figure that shows the minimum hardware emulation technique for a channel with multiple clusters

Fig. 5.6 shows a 3-taps channel emulated utilizing the cascaded technique. A pattern of 1 followed by a large number of 0s was sent by the generator with Amplitude Shift Keying (ASK) modulation. Note that such a channel is considered to be very frequency selective for the sent signal since the delay between the selected taps is too large as compared to the symbol duration where ISI takes place. However, the chosen symbol rate, and modulation scheme is for the sake of clarifying the channel delay taps on the oscilloscope where the taps are too obvious in the shown figure. The utilized 813-SL140.0M-77A bandpass SAW filter is shown in Fig. 5.7. It offers absolute group delay of 0.78 µsec at a center frequency of 140 MHz, and a group delay variation of up tp 40 nsec among a huge bandwidth of 70 MHz. The filter is characterized by a huge insertion loss of up to

Chapter 5. Hardware implementation 46

25 dB at its center frequency. Thus, an amplifier is attached with each SAW filter to compensate this significant loss.

Figure 5.6: A 3-taps channel emulated using two SAW filters

Figure 5.7: The utilized lossy SAW filter cascaded by an amplifier to make up for the huge insertion loss caused by the SAW filter

A valid option to increase the flexibility of the delay of the channel taps is by switching the signal between several parallel branches. Each branch is characterized by its delay. However, the offered flexibility is still much lower than the software emulation.

Chapter 5. Hardware implementation 47